The goal of the proposed study is to develop and prospectively validate the EMR Risk of Alzheimer's and Dementia Assessment Rule (eRADAR), a simple, automated tool that will use information in the electronic medical record (EMR) to identify patients likely to have current, undiagnosed dementia. There are currently more than 5 million people in the U.S. living with Alzheimer's disease and other dementias, and only half of them have been diagnosed. Lack of diagnosis in these patients can cause major problems for the healthcare system and society at large. Patients with undiagnosed dementia are more likely to need emergency care for comorbid conditions. They also are more likely to engage in risky behaviors such as driving. Current clinical guidelines recommend diagnosis of Alzheimer's and dementia as early as possible in the disease process to provide patients with optimal care. Clinicians can make adjustments to care plans, such as simplification of medication regimens and greater involvement of family members. Patients can be provided with an opportunity to participate in planning for their future care needs by communicating their goals and preferences and completing advance directives. Caregivers can receive earlier support, which can reduce stress and burnout. One of the key barriers to timely diagnosis of Alzheimer's disease and dementia is lack of time and resources for clinicians to identify high-risk patients. Many patients do not present with obvious symptoms of cognitive impairment, particularly early in the disease process, making these patients difficult to detect during brief clinical encounters. We propose to address this barrier to early diagnosis by developing and prospectively validating the EMR Risk of Alzheimer's and Dementia Assessment Rule (eRADAR), which will use information in the EMR (such as unfilled medications, clinic `no shows' and visits for dementia-related symptoms) to detect patients who are likely to have current, undiagnosed dementia. We will use a unique data source that combines `gold standard' dementia incidence data from the National Institute on Aging (NIA)- funded Adult Changes in Thought (ACT) study?a prospective cohort study of dementia in more than 5,000 older adults?with EMR data from the Group Health integrated healthcare delivery system. Our Specific Aims are: 1) to develop and internally validate eRADAR using dementia incidence data from ACT linked with EMR data from Group Health; 2) to craft optimal strategies for testing and implementing eRADAR, heavily informed by input from patients, caregivers, clinicians and key healthcare system stakeholders; and 3) to prospectively assess accuracy, clinical test characteristics, feasibility and acceptability of eRADAR in Group Health to inform a larger clinical trial of impact on patient care and clinical outcomes. Our ultimate goal is to improve quality of life in people living in the community with dementia through earlier detection and optimal care management. If eRADAR proves to be accurate and acceptable in this study, it could serve as a national model for timely diagnosis of Alzheimer's and dementia.